University of Edinburgh logo British Antarctic Survey logo National Oceanographic Centre logo University of Leeds logo

PhD Projects

Sensors and Satellites in the Almond: Water Quality Monitoring

Supervisors: Prof Lindsay Beevers(University of Edinburgh), Dr Encarni Medina-Lopez(University of Edinburgh), Professor Paul Kay, Dr Phil Taylor, Dr Peter Hunter, Dr Adrian Bas


For centuries, our riverine environments have been affected by shifting combinations of physical, biological and chemical stressors. Simultaneously, climate-change driven shifts in hydrological patterns and warming waters are expected to influence the function, physiology and phenology of freshwater ecological communities, with changing  exposure likely to magnify effects on biota (e.g. through changes in organism sensitivity). Pollutant cocktails from a range of sources impact our environment, and these climate induced changes to hydrological extremes are likely to continue these impacts and exacerbate them in the future.

A recently awarded NERC funded research project MOT4Rivers is developing monitoring systems in the Almond catchment (tributary to the River Forth in the central belt of Scotland) to explore water quality processes at a detailed catchment scale, and upscaling these findings to National Scale understanding. This system includes innovative sensor technologies for novel pollutants; thermal imaging of water flows using next generation high resolution satellite data; and eDNA sampling to enhance existing monitoring. The system will deliver high frequency measurements of pollutants, biota and ecologically-critical hydro-meteorological indicators across rural-urban transition zones.  This PhD would work with the research partners (led by University of Stirling) to explore the relationship between in-situ measurements, remote sensing and satellite data and novel sensors.

Data from the novel sensors must be ground truthed, and once these measurements are spatially and temporally matched across the sample sites there are open questions around whether relationships between satellite imagery and in-situ senor data exist. Inferring water quality measurements (e.g. novel pollutants) from satellite images is likely to be challenging as the relationship will be highly non-linear. Machine learning methods can be used to train for example, a deep neural network (DNN) to model this non-linear relationship directly from data. DNNs are flexible models and can be adapted to be multi-modal and could be trained on a collection of spatio-temporal data at the catchment scale. Additionally, in other parts of the MoT4Rivers project GANs (Generative Adversarial Network) will be explored for national scale relationships across different spatio-temporal water quality datasets, hence there is the opportunity to explore the work of this PhD could feed new and exciting data into the national scale research.

Study Site:

The river Almond is typical of a UK system that has been impacted by complex cocktails of pollutants, nested within the Forth catchment. The Almond is one of Scotland’s most polluted rivers, with significant shifts in economic development (shale mining to electronics and chemical industries; agricultural intensification; urbanisation) since the mid 19th century, resulting in a gradient of pollution hotspots and impairments. The MoT4Rivers project will set up three sentinel sites on the East Calder reaches of the Almond catchment and instrument rural-urban transition zones to monitor a range of land-use types, including legacy (e.g. mine waste) and contemporary (e.g. CSOs, health care facilities) pollution sources.

Aim of the PhD: Examine the output of novel river monitoring (range of sources described above), explore relationships between sensors, in-situ monitoring and remote sensing data in order to understand detailed river data and upscale for catchment scale water quality processes.

Key Research Questions:

  1. Can novel sensor data be ground truthed to in-situ measurements in the three sentinel sites in order to understand river water quality at high spatio-temporal resolution?
  2. Do relationships exist between the ground-truthed sensor data and remote sensing/satellite data that allows scaling up of the detailed river understanding to catchment scale or beyond?

Please be aware that due to funding requirements this project is only available to applicants from the UK or who have settled or pre-settled status in the UK  (i.e. a home fees student)